Structural and energetic basis for the molecular recognition of dual synthetic vs. natural inhibitors of EGFR/HER2.

Activation of EGFR starts by ligand binding at the extracellular domain which results in homo and heterodimerization, leading to phosphorylation, activation of downstream signaling pathways which upregulate expression of genes, proliferation and angiogenesis. Abnormalities in the expression of EGFR play a critical role in the development of different types of cancer. HER2 is the preferred heterodimerization partner for EGFR; this biological characteristic together with the high percentage of structural homology has been exploited in the design of dual synthetic inhibitors against EGFR/HER2. Herein we combined structural data and molecular dynamics (MD) simulations coupled to an MMGBSA approach to provide insight into the binding mechanism between two dual synthetics (lapatinib and TAK-285) and one dual natural inhibitor (EGCG) which target EGFR/HER2. In addition, we proposed some EGCG derivatives which were filtered through in silico screening. Structural analysis demonstrated that the coupling of synthetic, natural or newly designed compounds impacts the conformational space of EGFR and HER2 differently. Energetic analysis points out that lapatinib and TAK-285 have better affinity for inactive EGFR than the active EGFR state or HER2, whereas some EGCG derivatives seem to form binding affinities similar to those observed for lapatinib or TAK-285.

Binding free energy calculations using MMPB/GBSA approaches for PAMAM-G4-drug complexes at neutral, basic and acid pH conditions.

Dendrimers are synthetic macromolecules with a highly-branched structure and high concentration of surface groups. Among dendrimers, Poly(amidoamine) (PAMAM) has received substantial attention as a novel drug carrier and delivery system. Depending on the generation and type of terminal groups, dendrimer toxicity could change and include cytotoxicity. Although PAMAM is water soluble, molecular modeling of the dendrimer-drug complex is considered challenging for exploring the conformational mobility of dendrimers and atomic specific interactions during the dendrimer-drug association. However, conventional protocols for predicting binding affinities have been designed for small protein molecules or protein-protein complexes that can be applied to study the dendrimer-drug association. In this work, we performed docking calculations for a set of 94 previously reported compounds on PAMAM of fourth generation (G4-PAMAM) to select six compounds, cromoglicic acid (CRO) - a mast cell stabilizer, Fusidic acid (FUS) - a bacteriostatic antibiotic, and Methotrexate (MTX) - a chemotherapy agent and immune system suppressant, which have the highest affinities for G4-PAMAM, and Lidocaine (LDC) - used to numb tissue in a specific area and for ventricular tachycardia treatment, Metoprolol (MET) - a ß1 receptor blocker, and Pindolol (PIN) - a ß blocker, which have the lowest affinities for the G4-PAMAM dendrimer, to perform MD simulations combined with the molecular mechanics generalized/Poisson-Boltzmann surface area MMGBSA/MMPBSA approach to investigate the interactions of generating 4 charge-neutral, charge-basic and charge-acid G4-PAMAM dendrimers. In addition, to validate these theoretical G4-PAMAM-drug complexes, the complexes were experimentally conjugated to determine their stability in aqueous solubility studies immediately and over one year. Our results show that among the different commercial drugs, both charged and neutral PAMAM have the most favorable binding free energies for CRO, MTX, and FUS, which appears to be due to a complex counterbalance of electrostatics and van der Waals interactions. These theoretical and aqueous solubility studies supported the high affinity of methotrexate for the G4-PAMAM-drug due to its carboxyl and aryl moieties that favor its accommodation by noncovalent interactions.